INDEX

Explanations

mentions of the Premier League

oai_token-act-pair · gpt-3.5-turbo

references to the Premier League

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wal

-0.78

abeth

-0.77

raped

-0.74

raf

-0.69

ppo

-0.68

enegger

-0.66

 externalToEVAOnly

-0.66

ollow

-0.65

{"

-0.65

maxwell

-0.64

POSITIVE LOGITS

 League

1.31

League

0.97

 Division

0.86

 league

0.79

 Soccer

0.78

 Athletic

0.77

 Boxing

0.75

 Liga

0.72

 Development

0.72

idge

0.71

Activations Density 0.036%

mentions of the Premier League

references to the Premier League

No Comments

No Known Activations